AdCP and AAMP: The Two Agentic Ad Standards, and Which One a DTC Operator Should Track

If you have seen both acronyms and assumed one is the standard and the other is a product, here is the correction: AdCP and AAMP are two competing answers to the same question, and the bodies behind them are not collaborating. This is a standards contest, not a standard.

The question both are answering: when the buying and selling of advertising is done by software agents rather than by people clicking in a UI, what language do those agents speak to each other? The two camps disagree about whether that language should be new or should extend what already exists.

AdCP: build a new protocol for agents

AdCP, the Ad Context Protocol, is built on MCP and is usually described as OpenRTB for the AI era: agent-to-agent negotiation rather than impression-level auction. It is organised into four modules, Signals Activation for audience discovery, Media Buy for campaign execution, Creative for asset management, and a Curation module. FreeWheel and NBCUniversal ran the first MCP-based agent-to-agent media buy in January 2026.

It launched in October 2025 with a founding consortium, but describing it as any one company's protocol is now out of date. It is published by AgenticAdvertising.org, an independent non-profit, and the state of that organisation is worth stating precisely because it is easy to overstate.

The Rothenberg detail is the one that tells you how serious the contest is. The rival camp recruited the former chief of the standards body it is competing with, as an interim advisor.

AAMP: extend what the sell side already implements

AAMP, the Agentic Advertising Management Protocols, is the IAB Tech Lab's umbrella initiative. It was named on 26 February 2026 and published as a framework on 16 March 2026, and it reached version 2.3 by mid 2026.

It is organised in three pillars rather than four modules: Agentic Foundations, which is execution infrastructure including the Agentic Real Time Framework; Agentic Protocols, the management layer covering Agentic Direct, Agentic Audiences and Agentic Mobile; and Trust and Transparency, which is the Agent Registry providing verified agent identity and disclosure.

The strategic difference is not technical. AAMP extends standards the sell side has already implemented against for years. AdCP proposes new ones designed for agents from the start. Trade coverage of the dispute has been direct about the fact that the Tech Lab does not regard the two as compatible.

The comparison, such as it is

What a DTC operator should actually do

Mostly nothing, and it is worth being blunt about that rather than manufacturing urgency.

• Neither is something you adopt. These are protocols between buying and selling systems. Unless you buy premium video or CTV directly, you will encounter them through whatever your DSP or ad platform implements, on their timeline and not yours.
• Ask which component, never whether a vendor supports the standard. Both are umbrellas over parts at very different maturities. Agentic Audiences and the Agent Registry are different problems with different adoption curves, and a claim without a version and a component name is not a claim.
• The identity half is the part that touches you. The Agent Registry, and adjacent work like agent identity at the CDN layer, is what determines whether your bot mitigation rejects a legitimate shopping agent. That is a live operational question today, unlike the media-buying half.
• Do not pick a winner. The contest is unresolved, the bylaws of one side describe a board that has not been elected yet, and version numbers on the other are moving quarterly.

The media-buying half of agentic advertising is a standards fight between vendors. The identity half is already deciding whether an assistant can open your store.

That second half is not theoretical. Six stores in our public registry return 403 to every request, from a crawler and from a desktop browser user agent alike, which means an assistant cannot read them at all. No protocol resolves that. It is a rule at the edge, and it is the agent-commerce problem an operator can actually act on this quarter.

Both of these may matter enormously in two years. Neither changes what you should do next week. If you want the thing in agent commerce that is already measurable, it is whether assistants name your brand when somebody asks what to buy in your category, and that does not depend on either standard shipping.